[ICIP 2023] Exploring Diffusion Models For Unsupervised Video Anomaly Detection

Exploring Diffusion Models For Unsupervised Video Anomaly Detection
Anil Osman Tur, Nicola Dall'Asen, Cigdem Beyan, Elisa Ricci
University of Trento, Fondazione Bruno Kessler, Trento, Italy,

DOI: 10.1109/ICIP49359.2023.10222594

Installation

Please follow the instructions in INSTALL.md.

Dataset and Data Preparation

Please follow the instructions in DATASET.md for data preparation.

Diffusion Model

Implemented diffusion model is in the k_diffusion/models/feature_v1.py file. The model is trained with train_ano.py script.

Autoencoder Model

The autoencoder model is re-implemented from the descriptions of the paper Generative Cooperative Learning for Unsupervised Video Anomaly Detection. Used for generating the baselines for the paper.

Citation:

Please use the following BibTeX entry for citation.

@INPROCEEDINGS{tur2023exploring,
  author={Tur, Anil Osman and Dall’Asen, Nicola and Beyan, Cigdem and Ricci, Elisa},
  booktitle={2023 IEEE International Conference on Image Processing (ICIP)}, 
  title={Exploring Diffusion Models for Unsupervised Video Anomaly Detection}, 
  year={2023},
  volume={},
  number={},
  pages={2540-2544},
  doi={10.1109/ICIP49359.2023.10222594}}